Unverified50% confidenceFactExact time
Fixed an issue where MCP tool calls in restored sessions failed because the server was still connecting; calls now wait up to 10 seconds.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/28/2026
First Seen
Valid until: 12/27/2026
Sources
Related Entities
Related Claims
UnverifiedFixed an issue where SDK and cloud sessions would hang indefinitely when an MCP server handshake acknowledgment was lost; they now time out after 70 seconds and mark only that server as failed76% similarUnverifiedMCP fixes include: progress notifications for background tool calls are no longer dropped, transient HTTP 404s returned by stateless remote MCP servers no longer render the entire session unusable, and stdio MCP servers that have not finished starting up are now properly cleaned up when the session ends.74% similarUnverifiedThe update fixes an issue where the tool list cache was retained after an MCP server reconnected.73% similarUnverifiedA /mcp reconnect all command has been added for MCP, enabling a single retry of all MCP servers that have failed to connect or require authentication.72% similarUnverifiedv2.1.274 fixes an issue where Streamable HTTP MCP tool calls would time out after approximately 5 minutes even when a longer timeout was configured.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/957225API
curl https://kongchang.com/api/v1/knowledge/claims/957225MCP
get_claim(id=957225)